Cold email isn't dead. Emails about you are.

You already sent the batch. Nobody answered.

Opens were fine. They still deleted it. They can tell, in one second, that it could have gone to anyone. The next one has to carry a finding about their product, built by your tool, that they can check in thirty seconds. No finding, no send.

Campaigns are born paused. Nothing goes out until you have read the exact email and pressed send.

6,807 SAAS INDEXED  ·  11 LAUNCH SOURCES SWEPT EVERY MORNING

How it works

Three steps. No templates. One proof per email.

01

We find the launch

11 launch sources swept every morning, around 200 new products a day. The index holds 6,807 SaaS. We keep the ones that need what you sell, not the ones that compete with you.

02

Your endpoint builds the proof

We POST each matched lead to your app, HMAC signed, with its name, its URL and what it does. Your app returns the subject, the HTML and the text. No proof, no email.

03

You approve, then it sends

You read the exact email, byte for byte, before sending is on. Then it goes out with the PROOF INSIDE block, paced on a warm up curve, and you get opens, clicks and bounces lead by lead.

A letter from the founderAUGUST 2026

I sent thousands of cold emails. Almost nobody replied.

Proofdrop exists because of the handful that did reply, and the one thing they had in common.

I was selling a logging tool and I did everything the playbooks say. A clean list. A personalized first line for every prospect. Four-step sequences. Thousands of emails went out. A handful of people replied. Nobody bought.

I stopped because the numbers were embarrassing, not because I ran out of names. The list was still full.

Later I went back and read the replies again. They were all the same kind of email. They were the ones where I had done something first: signed up and walked through their onboarding, found a redirect that was broken, measured their API for an hour. Then I opened with what I had found instead of with what I was selling.

Each one of those took me about forty minutes, which is why I sent so few. So the question stopped being about copy and turned mechanical. What would it take to run the forty-minute version every morning, for every product that launched that day, with nobody sitting in the middle of it?

That is Proofdrop. You own the message and the evidence. We run the boring machinery that decides whether a good email ever arrives: sending, pacing, quotas, dedup, bounces, unsubscribes.

If they can tell it wasn't for them, they won't answer. The next email has to carry something they can check.

Carol Campistol

Founder · Proofdrop

support@mail.proofdrop.net

The artifact

This is the email. Every number in it can be checked in thirty seconds.

A monitoring tool writing to a founder who launched this morning. The evidence was generated against her own API, minutes before the email went out, so she never has to take the sender's word for anything.

Illustrative example · upcheck.io and meterbase.dev are not real companies
PROOF-0412.EMLSENT 09:12 · LAUNCH DAY +0
FROMDana Whitfield <dana@upcheck.io>
TOpriya@meterbase.dev
SUBJECT/v1/usage returned 502 on 3 of 20 probes this morning

Your /v1/usage endpoint was failing intermittently this morning. I run an uptime and API monitoring service. Meterbase hit Show HN at 06:00, so I pointed our probes at your public API and ran the check any developer runs in their first ten minutes.

PROOF INSIDEUPCHECK.IO/PROBE · 20 PROBES · 06:14 TO 07:49 · COMPOSED 09:07
06:14:02GET /v1/usage200212ms
06:19:02GET /v1/usage50248ms
06:24:02GET /v1/usage50251ms
06:29:02GET /v1/usage50247ms
06:34:02GET /v1/usage200208ms

3 FAILURES IN 20 PROBES · SHOWING 06:14 TO 06:34 · REQUEST IDS IN THE FULL LOG

Three failures in twenty probes, all from the same edge. Separately: your docs promise a 24h retry backoff on failed usage events, but the worker retries inside the minute. The same event id came back four times in five minutes.

Want the full probe log with request ids?

DANA WHITFIELD · UPCHECK.IO

97

SaaS added today

Pricing

One number changes between plans: how many go out a day.

Every plan runs the same machine. The 11 launch sources swept every morning, your own sending domain, the exact email in front of you before anything leaves, and the rule that a lead without a proof does not get written to. What you pick is the ceiling.

Starter

€19

per month, VAT on top

For finding out whether a proof gets replies where a pitch never did.

  • 25 proofs a day, per campaign
  • All 11 launch sources, swept every morning
  • Your own sending subdomain
  • The exact email before anything leaves
  • No proof, no send
  • Opens, clicks and bounces, lead by lead
  • One email per person. Ever
  • Unsubscribe on every email
Start with Starter

Campaign starts paused. Cancel any month.

RECOMMENDED

Operator

€49

per month, VAT on top

For making this the outbound channel, not an experiment beside it.

  • 50 proofs a day, per campaign
  • All 11 launch sources, swept every morning
  • Your own sending subdomain
  • The exact email before anything leaves
  • No proof, no send
  • Opens, clicks and bounces, lead by lead
  • One email per person. Ever
  • Unsubscribe on every email
Start with Operator

Campaign starts paused. Cancel any month.

Pro

€99

per month, VAT on top

For agencies and for anyone running outbound for several products at once.

  • 200 proofs a day, per campaign
  • Everything in Operator
Start with Pro

Campaign starts paused. Cancel any month.

Before you pay

Six things founders ask us first.

01Who actually writes the proof?

Your endpoint does. We POST each lead to your app, HMAC signed, with its name, its URL and what it does. Your app builds the proof and returns the subject, the HTML and the text.

Proofdrop sends that byte for byte. We do not rewrite it, shorten it or restyle it. The only thing we add is the delivery footer at the very end, because a cold email without a way out is illegal and burns the domain.

02Are you going to burn my domain?

You send from mail.yourdomain.com, a subdomain you set up with records we hand you. Never the root domain, so your company mail is never in the blast radius. If the DNS is not verified yet, nothing sends at all.

A new domain ramps on a fixed curve: 10 emails on day one, 60 by the end of week one, your full ceiling from week three. Emails leave between 09:00 and 23:00 in the timezone you set on the campaign, on the days you pick, with irregular gaps instead of one every five minutes on the dot.

03What happens if my endpoint breaks?

Nothing goes out. No proof, no email. That is the whole rule.

A lead whose generation fails is retried up to 3 times and then dropped, and a generator that keeps answering pending gets its slot released after 8 tries so the rest of the campaign keeps moving. Proofdrop never invents a proof to fill the hole, and it never falls back to a template with a first name at the top.

04Is this spam?

Every email carries a signed unsubscribe link plus the List-Unsubscribe headers that draw the native opt out button in Gmail and Outlook. It is not optional and you cannot switch it off.

One person gets one email, ever, even when two different SaaS share the same contact address. Anyone who replies is never written to again, from any campaign. An unsubscribe or a spam complaint removes that address globally, not just from your campaign, and we stop generating proofs for them before we spend a call on your API.

05Where do the leads come from?

11 launch directories and feeds swept every morning, around 200 new products a day. The index holds 6,807 SaaS right now.

Nothing enters without a description and a needs and provides classification, or it gets rejected at the door. We do not buy lists, and the addresses are not exported or resold. The matching looks for products that need what you sell, not products that compete with you.

06What if I do not have an endpoint yet?

You can write the email once inside the campaign and Proofdrop will send exactly that to every matched lead, with the same pacing, the same caps and the same tracking.

Be honest about what that is: a cold email with no proof inside it, which is the thing that was not working. It is a way to start while you build the endpoint, not the product.

Pull up the last email you sent. Take your name off it. Could anyone have sent it?

If yes, they already deleted it. The next one has to be about their product, not yours, with a finding they can verify in thirty seconds.

Every morning 11 sources fill with products that launched today. The index already holds 6,807 of them. Your endpoint builds the finding. You read the exact email. Then it leaves, or it doesn't.

Send an email they can check →

No card. Every campaign starts paused, so nothing leaves until you say so.